by Caroline Guillaume, CEO at TrustInSoft.
Industries are increasingly dependent on error-free software to ensure short-term success and long-term viability.
Although the number of applications for software analysis is growing exponentially as the need to ensure software safety and security expands in areas like aeronautics, defense, space, medical, consumer devices and many other industries, the one that is of the most immediate concern to many is software devoted to the automotive industry.
Why? Because most of us either drive cars, ride in them as passengers, or hire them in the form of rental cars, taxis, buses, or other types of transport link. The veracity of that software is arguably just as important to human safety and business longevity as it is to the software that helps control planes, trains, and spacecraft.
A good example is the traditional London black taxi, which remains black but with a vanishing few emitting the once familiar clattering sound of a diesel engine. That’s because, since “London Electric” vehicles began making trial appearances on the capitol’s boroughs, districts, neighborhoods, and villages in 2015, according to Transport of London statistics published in December 2023, they now make up more than 50 per cent of London’s nearly 14,700 black taxis. These “zero emission capable” (ZEC) taxis mean they are either fully electric or hydrogen fuel-cell driven, which also means they are increasingly reliant on software to maintain and control their functionality.
The focus of this article, by virtue of the sheer numbers involved in the transition to electric, is software safety and security for the automotive industry. As the automotive industry continues to embrace digital transformation, the complexity of embedded software in vehicles—irrespective of the propulsion system—is gaining serious traction. And it’s that pace of need that has become a challenge.
Everything from advanced driver assistance systems (ADAS) to autonomous driving features require error, i.e., “bug”-free operational, monitoring, and control software to ensure passenger and pedestrian safety, vehicle reliability, compliance with government standards, and the ability to meet or exceed industry performance expectations. That requires rigorous, time-consuming, and often expensive software testing. Testing needs to be done increasingly quickly and accurately.
That’s because automotive software operates in safety-critical systems where even minor errors can cause catastrophic consequences. Malfunctions in ADAS, braking systems, or engine control units (ECUs) can result in accidents, injuries, or even loss of life. And because the automotive industry quite rightly adheres to rigorous safety standards such as ISO 26262, error detection ensures compliance with these standards, allowing manufacturers to demonstrate due diligence in safety assurance.
This is more important than ever because modern vehicles consist of thousands of interconnected ECUs, sensors, actuators, and communication networks. Detecting software errors across this intricate ecosystem is challenging, but essential.
And not just for safety reasons. Early detection of errors can save manufacturers huge amounts of money and reputational damage because they can identify, intercept, and rectify errors in proposed software code and ensure those errors are eliminated during the development phase, which is significantly more cost-effective than addressing them post-manufacturing or in the field. Advanced error detectionsubstantially reduces recalls and warranty claims for the manufacturer and significantly reduces the headache of maintenance costs for the buyer.
In the automotive industry, software error detection is not a “nice-to- have”—it’s a necessity. The power of TrustInSoft Analyzer lies in its ability to enable automotive software developers to create safer, more reliable systems by using formal methods to ensure code correctness with 100% mathematically proven certainty.
And now a word about how TrustInSoft intertwines with AUTOSAR (AUTomotive Open System Architecture, formerly known as AUTOSAR R3.x). AUTOSAR is widely used in the automotive industry for developing embedded software. It provides
a standardized framework for ECUs, communication stacks, and software components.
Where TrustInSoft Analyzer comes in is that it verifies AUTOSAR Classic- compliant code for safety and correctness. It ensures that software components adhere to AUTOSAR guidelines while avoiding run time errors and memory related issues common to the C language. TrustInSoft Analyzer helps provide a high level of code coverage and robust analysis while avoiding pitfalls commonly associated with other static analyzers. It does this by leveraging formal methods combined with full path and context sensitivity of the code analyzed and the ARXML file format itself. It ensures that input coverage testing is exhaustive while also adhering to the boundary conditions defined within. Thanks to the accuracy of the formal methods, it avoids false positives as well as ensuring that the generalization, which tests all possible inputs, adheres to the boundary conditions and other requirements outlined in the ARXML.
Because it is itself a highly modular tool, TrustInSoft Analyzer integrates seamlessly with AUTOSAR tool chains, enabling automotive software developers to incorporate formal verification into their existing AUTOSAR workflows. Formal methods-driven static code analysis and TrustInSoft have long been associated with the highest levels of formal verification for the most safety critical use cases within the automotive industry. Moreover, because TrustInSoft Analyzer is ISO 26262 qualified, it can assist in achieving ISO 26262 certification by providing unassailable evidence of code correctness.
Since AUTOSAR Classic remains a predominant software development environment, TrustInSoft Analyzer’s easy compatibility ensures that error detection tracks seamlessly with industry standards and won’t delay ongoing development efforts. In fact, the reality is that TrustInSoft Analyzer can accelerate development efforts and testing practices. And, as we’ve already seen, the demand for ever- improving software is increasing rapidly.
However, that increase in demand is not limited to the automotive industry. As hugely important as this is, it is only one of scores of industries that are increasingly dependent on error-free software to ensure their short-term success and long-term viability.
Every line of software code matters— especially when lives are at stake. TrustInSoft Analyzer helps ensure that those lines of code are error-free and dependable. By adopting TrustInSoft Analyzer, automotive manufacturers and suppliers can enhance safety, reduce costs, and build trust in their software-driven vehicles.
The software in London’s electric black cabs means they are now much cleaner, safer, and more reliable, and there is precious little in the world that is more dependable than a London taxi.